MEMORANDUM

DATE:09/19/2014
FROM: 		Laura Farris/EPA Region 8
TO:CLEAN Power Plan for Existing Power Plants; Docket Id: OAR - 2013-0602
SUBJECT:	Montana PSC Conference Call held on 09/19/2014

SUMMARY:        
At the request of Travis Kavulla, Commissioner, Montana Public Service Commission (PSC), staff from EPA Headquarters, EPA Region 8, and EPA Region 9 participated in a conference call with the Montana PSC and others on EPA's use of the integrated planning model (IPM) for the Clean Power Plan on 09/19/2014. The Clean Power Plan was proposed on June 2, 2014.

ATTENDEES:

EPA Headquarters
Gabrielle Stevens, Bill Meroney, Brian Fisher, Reid Harvey, Mikhail Adamantiades, Jeb Stenhouse, Jeremy Mark

EPA Region 8
Laura Farris

EPA Region 9 
Ben Machol

External Stakeholders
Commissioner, Travis Kavulla, Margo Schurman and Bob Decker, Montana PSC
Peter Ashcroft, Utah Office of Energy Development
Keegan Moyer, Western Electricity Coordinating Council
Tom Carr and Alaine Ginocchio, Western Interstate Energy Board
Karen Griffin, Grace Anderson, and Angela Tanghetti, California Energy Commission
Ed Stoneburg, Arizona Corporation Commission

Questions Discussed
(1) We've heard from RTOs some criticism of the deficiency of IPM as a hub-and-spoke model that doesn't accurately represent regions but instead uses NERC regions, which do not represent current market footprints. Could IPM explain how it defines regions and uses them in its model?  
(2) A fuller explanation of what it is meant to be a 'hub-and-spoke' model 
(3) A description of how transmission is modeled, or could be modeled/assumed as a supplement to modeling results. (On this point, thinking of the utilization, load-duration, and power flow studies that certain NERC designated Regional Entities engage in.) 
(4) A description of the source data for cost analysis. 
(5) What is the time-period granularity of the model (some software represents dispatch on an hourly basis, others on an intra-hourly basis) 
(6) How does the model incorporate, if at all, stochasticity and the quantification of uncertainty as users are asked to feed in assumptions to populate the model. 



